Fig. 4
From: Enhancing image resolution of confocal fluorescence microscopy with deep learning
Super-resolution imaging of the microtubules using TCAN. (a) The first column of images is diffraction-limited confocal microscopy images of the microtubules used as input to the network. (b) The second column of images is the super-resolved output images. (c) The third column of images is the STED images of the same field of view. (d) The fourth column of images is the deconvolution of the STED images of the same field of view. The green box regions are shown below at a magnified scale. (e-g) Statistical comparison of confocal images, output images, STED images and deconvolution of STED images in terms of MSE, SNR and resolution. Tukey box-and-whisker plots with outliers displayed as diamonds are shown. Experiments are repeated with 15 images, achieving similar results. Scale bars in the second row, the fourth row, the sixth row, the seventh row and the eighth row are 0.75 μm, 0.75 μm, 0.5 μm, 3 μm and 0.5 μm, respectively